ucscxenatoolspy

Python port of the R package UCSCXenaTools — download and explore datasets from UCSC Xena data hubs.

Installation

From PyPI (recommended):

pip install ucscxenatoolspy

From local source:

# Clone or download the source code first
pip install .              # standard install
pip install -e .           # editable install (for development)

Quick Start

Workflow (Generate → Filter → Query → Download → Prepare)

from ucscxenatoolspy import xena_generate, xena_filter, xena_query, xena_download, xena_prepare

# Generate: select datasets from metadata
hub = xena_generate(subset=lambda df: df["XenaHostNames"] == "tcgaHub")
# class: XenaHub
# hosts():
#   https://tcga.xenahubs.net
# cohorts() (38 total):
#   TCGA Ovarian Cancer (OV), TCGA Kidney Clear Cell Carcinoma (KIRC), ...
# datasets() (715 total):
#   TCGA.OV.sampleMap/HumanMethylation27, ...  (715 total)

# Filter: refine by patterns
hub = xena_filter(hub, filter_datasets="LUSC_clinicalMatrix")
# class: XenaHub
# hosts():
#   https://tcga.xenahubs.net
# cohorts() (1 total):
#   TCGA Lung Squamous Cell Carcinoma (LUSC)
# datasets() (1 total):
#   TCGA.LUSC.sampleMap/LUSC_clinicalMatrix

# Query: build download URLs
query = xena_query(hub)
# This will check url status, please be patient.
# QueryResult(hosts=['https://tcga.xenahubs.net'], datasets=['TCGA.LUSC.sampleMap/LUSC_clinicalMatrix'],
#             urls=['https://tcga.xenahubs.net/download/TCGA.LUSC.sampleMap/LUSC_clinicalMatrix'], destfiles=[])

# Download: fetch files
result = xena_download(query)
# All downloaded files will be under: C:\Users\xxx\AppData\Local\Temp

# Prepare: load into pandas
data = xena_prepare(result)

Fetch API (Direct queries without downloading)

from ucscxenatoolspy import (
    fetch_dataset_samples,
    fetch_dataset_identifiers,
    fetch_dense_values,
    fetch_sparse_values,
    has_probeMap,
)

host = "https://toil.xenahubs.net"
dataset = "tcga_RSEM_gene_tpm"

# Get sample IDs
samples = fetch_dataset_samples(host, dataset, limit=10)
# ['TCGA-19-1787-01', 'TCGA-S9-A7J2-01', 'TCGA-G3-A3CH-11', 'TCGA-EK-A2RE-01',
#  'TCGA-44-6778-01', 'TCGA-F4-6854-01', 'TCGA-AB-2863-03', 'TCGA-C8-A1HL-01',
#  'TCGA-EW-A2FS-01', 'TCGA-IR-A3L7-01']

# Get gene/probe identifiers
identifiers = fetch_dataset_identifiers(host, dataset)
# ['ENSG00000000003.14', 'ENSG00000000005.5', 'ENSG00000000419.12',
#  'ENSG00000000457.13', 'ENSG00000000460.16', ...]

# Fetch expression values (with gene symbol via probeMap)
values = fetch_dense_values(
    host, dataset,
    identifiers=["TP53", "BRCA1"],
    samples=samples[:5],
    use_probeMap=True
)
#       TCGA-19-1787-01  TCGA-S9-A7J2-01  TCGA-G3-A3CH-11  TCGA-EK-A2RE-01  TCGA-44-6778-01
# TP53            5.887            5.517            2.382            4.591            5.299
# BRCA1           3.475            1.930            0.670            3.392            2.674

# Check if dataset has probeMap
if has_probeMap(host, dataset):
    print("Gene symbol queries available")
# Gene symbol queries available

Query Molecule Value of Selected dataset (Recommended)

from ucscxenatoolspy import query_molecule_value

# Single gene — returns pd.Series indexed by sample
values = query_molecule_value(
    "tcga_RSEM_gene_tpm",
    "TP53",
)
# TCGA-19-1787-01    5.887
# TCGA-S9-A7J2-01    5.517
# TCGA-G3-A3CH-11    2.382
# TCGA-EK-A2RE-01    4.591
# TCGA-44-6778-01    5.299
# TCGA-F4-6854-01    6.864
# Name: TP53, Length: 10535, dtype: float64

# Genomic signature formula — auto-queries each gene and evaluates
signature = query_molecule_value(
    "tcga_RSEM_gene_tpm",
    "TP53 + KRAS",
)
# Querying multiple identifiers: TP53, KRAS
# TCGA-19-1787-01     9.838
# TCGA-S9-A7J2-01     9.501
# TCGA-G3-A3CH-11     4.198
# TCGA-EK-A2RE-01     8.202
# TCGA-44-6778-01     9.363
# TCGA-F4-6854-01    10.254
# Name: TP53 + KRAS, Length: 10535, dtype: float64

Datsets Metadata (See documentation for more details)

from ucscxenatoolspy import load_xena_data, xena_data_update

# Load bundled metadata snapshot
df = load_xena_data()
# shape: (2314, 17)
# columns: XenaHosts, XenaHostNames, XenaCohorts, XenaDatasets,
#          SampleCount, DataSubtype, Label, Type, AnatomicalOrigin,
#          SampleType, Tags, ProbeMap, LongTitle, Citation, Version, Unit, Platform

# Fetch fresh metadata from all hubs
df = xena_data_update()
# Fetching metadata from UCSC Xena hubs...
#   [1/12] Querying tcgaHub... 715 datasets
#   [2/12] Querying icgcHub... ...
#   ...
# Total: 2314 datasets from 12 hosts

Built-in general phenotype for TCGA samples

from ucscxenatoolspy import tcga_clinical, tcga_survival

# Clinical annotations (Sample, Cancer, Age, Stage, Grade, etc.)
clinical = tcga_clinical()
#           Sample Cancer   Age Code  Gender Stage_ajcc Stage_clinical Grade
# 0  TCGA-OR-A5J1-01    ACC  58.0   TP    MALE   Stage II            NaN   NaN
# 1  TCGA-OR-A5J2-01    ACC  44.0   TP  FEMALE   Stage IV            NaN   NaN
# 2  TCGA-OR-A5J3-01    ACC  23.0   TP  FEMALE  Stage III            NaN   NaN
# 3  TCGA-OR-A5J4-01    ACC  23.0   TP  FEMALE   Stage IV            NaN   NaN
# 4  TCGA-OR-A5J5-01    ACC  30.0   TP    MALE  Stage III            NaN   NaN
# 5  TCGA-OR-A5J6-01    ACC  29.0   TP  FEMALE   Stage II            NaN   NaN

# Survival endpoints (OS, DSS, DFI, PFI with time)
survival = tcga_survival()
#           Sample   OS  OS.time  DSS  DSS.time  DFI  DFI.time  PFI  PFI.time
# 0  TCGA-OR-A5J1-01  1.0   1355.0  1.0    1355.0  1.0     754.0  1.0     754.0
# 1  TCGA-OR-A5J2-01  1.0   1677.0  1.0    1677.0  NaN       NaN  1.0     289.0
# 2  TCGA-OR-A5J3-01  0.0   2091.0  0.0    2091.0  1.0      53.0  1.0      53.0
# 3  TCGA-OR-A5J5-01  1.0    365.0  1.0     365.0  NaN       NaN  1.0      50.0
# 4  TCGA-OR-A5J6-01  0.0   2703.0  0.0    2703.0  0.0    2703.0  0.0    2703.0
# 5  TCGA-OR-A5J7-01  1.0    490.0  1.0     490.0  NaN       NaN  1.0     162.0

TCGA Analysis API

A FastAPI web service providing statistical analysis on TCGA gene expression data. Start the server and query via browser or curl.

Live Demo: Two deployed instances are available — try the fast one first:

⚡ Primary (fast) — http://biotree.top:38123/ucscxena/ (always-on, ~0.2s response):
curl http://biotree.top:38123/ucscxena/health
curl "http://biotree.top:38123/ucscxena/api/v1/diff-expr?gene=TP53&cancer=LUAD"
🐢 Fallback — https://ucscxenatoolspy.onrender.com (Render free tier, may need ~30s cold start):
curl https://ucscxenatoolspy.onrender.com/health
curl "https://ucscxenatoolspy.onrender.com/api/v1/diff-expr?gene=TP53&cancer=LUAD"

Install the optional API dependencies before running the service:

pip install "ucscxenatoolspy[api]"

# or, from a local checkout
pip install -e ".[api]"

Start the server

cd ucscxenatoolspy
uvicorn ucscxenatoolspy.api_service.main:app --reload --port 8765

Interactive docs at http://127.0.0.1:8765/docs.

Public deployment settings

For an internet-facing deployment, configure these environment variables before starting uvicorn:

# Require clients to send X-API-Key: <one-of-these-values>
UCSCXENA_API_KEYS="change-me,second-key"

# Restrict browser callers to your frontend domains
UCSCXENA_CORS_ORIGINS="https://example.org,https://app.example.org"

# Disable interactive docs on the public host
UCSCXENA_ENABLE_DOCS=false

# If running behind a trusted reverse proxy, use X-Forwarded-For / X-Real-IP
UCSCXENA_TRUST_PROXY_HEADERS=true

# Use shared rate-limit state across workers/containers
UCSCXENA_RATE_LIMIT_STORAGE_URI="redis://localhost:6379/0"

# Bound expensive analysis work
UCSCXENA_ANALYSIS_CONCURRENCY=4
UCSCXENA_ANALYSIS_TIMEOUT_SECONDS=120

# Optional: disable external mygene fallback for unknown identifiers
UCSCXENA_ENABLE_MYGENE=false

Deployment guide

The API is a normal ASGI application:

ucscxenatoolspy.api_service.main:app

A minimal production deployment usually has uvicorn running on localhost and a reverse proxy such as Nginx or Caddy handling HTTPS, public routing, and request size/time limits.

Install the package on the server.

git clone <your-repo-url>
cd ucscxenatoolspy
python -m venv .venv
source .venv/bin/activate
python -m pip install -U pip
python -m pip install -e ".[api]"

Create an environment file for the service, for example /etc/ucscxena-api.env.

UCSCXENA_API_KEYS=replace-with-a-long-random-key
UCSCXENA_CORS_ORIGINS=https://your-frontend.example.org
UCSCXENA_ENABLE_DOCS=false
UCSCXENA_TRUST_PROXY_HEADERS=true
UCSCXENA_ANALYSIS_CONCURRENCY=4
UCSCXENA_ANALYSIS_TIMEOUT_SECONDS=120
UCSCXENA_RATE_LIMIT_STORAGE_URI=redis://127.0.0.1:6379/0

UCSCXENA_API_KEYS is optional for private/internal deployments, but recommended for any public endpoint. If you run multiple workers, containers, or servers, configure UCSCXENA_RATE_LIMIT_STORAGE_URI so rate limits are shared; otherwise slowapi falls back to in-memory state per process.

Start the API locally.

uvicorn ucscxenatoolspy.api_service.main:app \
  --host 127.0.0.1 \
  --port 8765 \
  --workers 1 \
  --proxy-headers \
  --forwarded-allow-ips 127.0.0.1

Keep --workers 1 unless you have configured Redis-backed rate limiting and have enough memory for repeated TCGA analysis workloads. Increase UCSCXENA_ANALYSIS_CONCURRENCY before increasing ASGI workers when you mainly need more parallel analysis capacity.

Manage the process with systemd.

[Unit]
Description=ucscxenatoolspy TCGA Analysis API
After=network.target

[Service]
User=www-data
Group=www-data
WorkingDirectory=/opt/ucscxenatoolspy
EnvironmentFile=/etc/ucscxena-api.env
ExecStart=/opt/ucscxenatoolspy/.venv/bin/uvicorn ucscxenatoolspy.api_service.main:app --host 127.0.0.1 --port 8765 --workers 1 --proxy-headers --forwarded-allow-ips 127.0.0.1
Restart=always
RestartSec=5

[Install]
WantedBy=multi-user.target

After saving it as /etc/systemd/system/ucscxena-api.service:

sudo systemctl daemon-reload
sudo systemctl enable --now ucscxena-api
sudo systemctl status ucscxena-api

Put Nginx in front of the local service.

server {
    listen 80;
    server_name api.example.org;

    location / {
        proxy_pass http://127.0.0.1:8765;
        proxy_http_version 1.1;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_set_header X-Forwarded-Proto $scheme;
        proxy_read_timeout 180s;
    }
}

Enable HTTPS with your normal certificate workflow, for example certbot or your platform load balancer. The FastAPI app should stay bound to 127.0.0.1 unless you intentionally expose it directly.

Verify the deployment.

curl https://api.example.org/health
curl -H "X-API-Key: replace-with-a-long-random-key" \
  "https://api.example.org/api/v1/diff-expr?gene=TP53&cancer=LUAD"

The expected health response is {"status":"ok"}. If docs are enabled for a non-public staging instance, the OpenAPI UI is available at /docs.

Endpoints

Method	Path	Rate	Description
GET	`/health`	none	Health check for monitoring probes
GET	`/`	60/min	API info and endpoint listing
GET	`/api/v1/cancers`	60/min	List all 33+ cancers with tumor/normal sample counts
GET	`/api/v1/diff-expr`	10/min	Differential expression: tumor vs normal (Wilcoxon)
GET	`/api/v1/corr`	10/min	Spearman correlation between two genes in tumors
GET	`/api/v1/survival`	5/min	Survival analysis: gene expression vs 4 endpoints (log-rank)

Legacy /api/* paths redirect to /api/v1/* with a 308 Permanent Redirect.

Differential Expression

GET /api/v1/diff-expr?gene=TP53&cancer=LUAD

{
  "gene": "TP53",
  "cancer": "LUAD",
  "cancer_full_name": "Lung adenocarcinoma",
  "dataset": "tcga_RSEM_gene_tpm",
  "tumor":   {"n": 501, "mean": 5.887, "median": 5.799},
  "normal":  {"n": 119, "mean": 4.123, "median": 4.001},
  "log2_fold_change": 1.764,
  "p_value": 2.3e-15,
  "test": "Wilcoxon rank-sum (Mann-Whitney U, two-sided)"
}

Parameters:

gene - HUGO gene symbol or supported identifier (TP53, EGFR, HER2, 7157, ENSG00000141510). Aliases and IDs are resolved with mygene and must map to a gene present in tcga_RSEM_gene_tpm.
cancer - TCGA abbreviation, case-insensitive (LUAD, brca, KIRC, ...). Requires at least 3 normal samples (e.g. LAML returns 400).

log2_fold_change is computed as the tumor mean minus the normal mean because tcga_RSEM_gene_tpm values are already log2(TPM + 0.001) transformed.

Gene Correlation

GET /api/v1/corr?gene1=TP53&gene2=EGFR&cancer=LUAD

{
  "gene1": "TP53",
  "gene2": "EGFR",
  "cancer": "LUAD",
  "cancer_full_name": "Lung adenocarcinoma",
  "sample_type": "primary_tumor",
  "n": 501,
  "spearman_r": 0.1234,
  "p_value": 0.0056,
  "test": "Spearman rank correlation (two-sided)"
}

Only primary tumor (TP) samples are used. Requires at least 5 common samples after intersecting both genes' expression data.

Survival Analysis

GET /api/v1/survival?gene=TP53&cancer=LUAD

{
  "gene": "TP53",
  "cancer": "LUAD",
  "cancer_full_name": "Lung adenocarcinoma",
  "sample_type": "primary_tumor",
  "test": "Log-rank test",
  "survival": {
    "OS": {
      "name": "Overall Survival",
      "n_total": 501, "n_events": 187,
      "median_cutoff": {
        "method": "median",
        "cutoff": 5.799,
        "high": {"n": 251, "n_events": 80, "mean_survival_days": 1200.5},
        "low":  {"n": 250, "n_events": 107, "mean_survival_days": 900.3},
        "p_value": 0.012
      },
      "optimal_cutoff": {
        "method": "optimal_cutoff",
        "cutoff": 6.234,
        "high": {"n": 150, "n_events": 35, "mean_survival_days": 1500.2},
        "low":  {"n": 351, "n_events": 152, "mean_survival_days": 800.1},
        "p_value": 0.0003
      },
      "optimal_cutoff_note": "Exploratory minimum-p scan across candidate cutoffs; p_value is not adjusted for multiple cutoff testing."
    },
    "DSS": { ... },
    "DFI": { ... },
    "PFI": { ... }
  }
}

Four survival endpoints: OS (Overall), DSS (Disease-Specific), DFI (Disease-Free Interval), PFI (Progression-Free Interval). Each evaluated with median cutoff (high vs low expression split at median) and optimal cutoff (scanned across p25-p75 for the smallest log-rank p value, minimum 10% in each group). The optimal-cutoff p value is exploratory and is not adjusted for testing multiple cutoffs. Requires at least 10 samples with both expression and survival data.

Gene Alias Resolution

The API accepts multiple gene identifier formats and resolves them automatically:

Input type	Example	Resolution
Standard symbol	`TP53`	Exact match (instant)
Common alias	`HER2`	mygene -> `ERBB2`
Entrez Gene ID	`7157`	mygene -> `TP53`
Ensembl ID	`ENSG00000141510`	mygene -> `TP53`
Case-insensitive	`tp53`	Case-insensitive lookup -> `TP53`

When a gene is resolved (not an exact match), the response includes gene_input with the original query string.

Architecture

src/ucscxenatoolspy/api_service/
  main.py        # FastAPI app, routes, middleware, rate limiting
  analysis.py    # Core stats: Wilcoxon, Spearman, log-rank
  gene_utils.py  # Gene name resolution (exact -> case-insensitive -> mygene)
  cache_utils.py # In-memory TTL cache with LRU eviction
  gene_list.txt  # Bundled valid gene list from tcga_RSEM_gene_tpm

Security: gene input is length-bounded and limited to gene identifier characters (^[A-Za-z0-9][A-Za-z0-9@._-]{0,29}$), rate limiting per endpoint, error messages sanitized (URLs and retry info stripped), X-Request-ID on every response.

Quick Test

# List available cancers
curl http://127.0.0.1:8765/api/v1/cancers | python -m json.tool

# Differential expression
curl "http://127.0.0.1:8765/api/v1/diff-expr?gene=TP53&cancer=LUAD" | python -m json.tool

# Gene correlation
curl "http://127.0.0.1:8765/api/v1/corr?gene1=TP53&gene2=EGFR&cancer=LUAD" | python -m json.tool

# Survival analysis
curl "http://127.0.0.1:8765/api/v1/survival?gene=TP53&cancer=LUAD" | python -m json.tool

# Health check
curl http://127.0.0.1:8765/health

AI Assistant Skills

Pre-built integrations that let AI coding assistants (Claude Code, OpenAI Codex, Cursor) query the TCGA Analysis API directly — ask about gene expression, correlation, or survival in natural language (English or Chinese), and the assistant answers with data from the API.

Supported Platforms

Platform	Directory	Type
Claude Code	`skills/claude/xena-tcga-gene-query/`	Skill (SKILL.md + helper script)
OpenAI Codex	`skills/codex/xena-tcga-gene-query/`	Agent (agent config + helper script)
Cursor	`skills/cursor/xena-tcga-gene-query/`	Rule (.mdc + helper script)

What They Do

Each skill/agent/rule teaches the AI assistant to:

List cancers — what TCGA cancers are available, which have normal tissue controls
Differential expression — tumor vs normal comparison for a gene (Wilcoxon test)
Gene correlation — Spearman correlation between two genes in primary tumors
Survival analysis — expression-associated survival across OS, DSS, DFI, PFI (log-rank)
Cancer name mapping — resolves Chinese/English common names (肺癌→LUAD+LUSC, 乳腺癌→BRCA, etc.) and gene aliases (HER2→ERBB2)

The assistant queries the deployed API at http://biotree.top:38123/ucscxena/ by default (fast, always-on), falling back to https://ucscxenatoolspy.onrender.com if needed — no local server setup is required.

Install a Skill

Claude Code — register the skill globally or per-project:

# Per-project (from repo root)
cp -r skills/claude/xena-tcga-gene-query .claude/skills/

# Or install into your user skills directory
cp -r skills/claude/xena-tcga-gene-query ~/.claude/skills/

Then restart Claude Code and ask: "Is TP53 upregulated in LUAD?"

Cursor — copy the rule into your project:

cp skills/cursor/xena-tcga-gene-query/xena-tcga-gene-query.mdc .cursor/rules/

The rule is set to alwaysApply: true, so it activates automatically in every chat.

OpenAI Codex — register as a custom agent:

Place the skills/codex/xena-tcga-gene-query/ directory in your Codex agents directory, or use the Codex dashboard to create an agent from agents/openai.yaml.

Example Queries

Once the skill is installed, you can ask questions like:

English	Chinese
"Is TP53 upregulated in LUAD?"	"TP53在肺癌中高表达吗？"
"Are EGFR and KRAS co-expressed in lung cancer?"	"EGFR和KRAS在肺癌中共表达吗？"
"Does HER2 expression affect breast cancer survival?"	"HER2高表达影响乳腺癌预后吗？"
"What cancers have normal tissue controls?"	"哪些癌症有正常组织对照？"
"Show me correlation between TP53 and MDM2 in GBM"	"TP53和MDM2在胶质母细胞瘤中的相关性"

The assistant will call the API endpoints and synthesize results with proper scientific framing (sample sizes, p-values, exploratory vs confirmatory caveats).

Features

Core workflow: Generate, Filter, Query, Download, Prepare (mirrors R package)
Fetch API: Direct API queries for samples, identifiers, dense/sparse values
Query Molecule Value: Single gene lookup and genomic signature formula evaluation
Built-in TCGA data: Clinical annotations and survival endpoints included out of the box
File caching: Persistent disk cache for query results (configurable via UCSCXENA_CACHE_DIR)
ProbeMap support: Query by gene symbols (via probe-to-gene mapping)
Metadata management: Bundled snapshot + live update from 12 Xena hubs
R-compatible naming: camelCase parameter names for easy migration

API Reference

Function	Description
`xena_generate(subset)`	Select datasets from metadata
`xena_filter(hub, filter_cohorts, filter_datasets)`	Filter by regex patterns
`xena_query(hub)`	Build download URLs, validate with HEAD request
`xena_download(query)`	Download files with progress bar
`xena_prepare(result)`	Load TSV files into pandas DataFrame
`load_xena_data()`	Load bundled XenaData metadata
`xena_data_update()`	Fetch fresh metadata from all hubs
`fetch_dataset_samples(host, dataset)`	Get sample IDs
`fetch_dataset_identifiers(host, dataset)`	Get probe/gene identifiers
`fetch_dense_values(host, dataset, ...)`	Query dense matrix data
`fetch_sparse_values(host, dataset, ...)`	Query sparse/mutation data
`has_probeMap(host, dataset)`	Check probeMap availability
`query_molecule_value(dataset, molecule, host)`	Single gene or formula query (returns pd.Series)
`get_data(dataset, identifier, host)`	Fetch single identifier with caching
`tcga_clinical()`	Load built-in TCGA clinical annotations
`tcga_survival()`	Load built-in TCGA survival endpoints

Supported Hubs

12 UCSC Xena hubs are supported by default:

Hub Name	URL
tcgaHub	https://tcga.xenahubs.net
icgcHub	https://icgc.xenahubs.net
toilHub	https://toil.xenahubs.net
pancanAtlasHub	https://pancanatlas.xenahubs.net
gdcHub	https://gdc.xenahubs.net
ccleHub	https://ccle.xenahubs.net
gtexHub	https://gtex.xenahubs.net
atlasHub	https://atlas.xenahubs.net
xenaHub	https://xena.xenahubs.net
publicHub	https://public.xenahubs.net
pcaHub	https://pca.xenahubs.net
tdiHub	https://tdi.xenahubs.net

License

GPL-3.0 (same as original R package)

Name		Name	Last commit message	Last commit date
Latest commit History 13 Commits
docs		docs
skills		skills
src/ucscxenatoolspy		src/ucscxenatoolspy
tests		tests
.gitignore		.gitignore
.readthedocs.yaml		.readthedocs.yaml
COMPARISON.md		COMPARISON.md
LICENSE		LICENSE
README.md		README.md
pyproject.toml		pyproject.toml

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

ucscxenatoolspy

Installation

Quick Start

Workflow (Generate → Filter → Query → Download → Prepare)

Fetch API (Direct queries without downloading)

Query Molecule Value of Selected dataset (Recommended)

Datsets Metadata (See documentation for more details)

Built-in general phenotype for TCGA samples

TCGA Analysis API

Start the server

Public deployment settings

Deployment guide

Endpoints

Differential Expression

Gene Correlation

Survival Analysis

Gene Alias Resolution

Architecture

Quick Test

AI Assistant Skills

Supported Platforms

What They Do

Install a Skill

Example Queries

Features

API Reference

Supported Hubs

License

About

Releases

Packages

Contributors

Languages

Folders and files

Latest commit

History

Repository files navigation

ucscxenatoolspy

Installation

Quick Start

Workflow (Generate → Filter → Query → Download → Prepare)

Fetch API (Direct queries without downloading)

Query Molecule Value of Selected dataset (Recommended)

Datsets Metadata (See documentation for more details)

Built-in general phenotype for TCGA samples

TCGA Analysis API

Start the server

Public deployment settings

Deployment guide

Endpoints

Differential Expression

Gene Correlation

Survival Analysis

Gene Alias Resolution

Architecture

Quick Test

AI Assistant Skills

Supported Platforms

What They Do

Install a Skill

Example Queries

Features

API Reference

Supported Hubs

License

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages